qestion for jncuse - q1&2 wins vs numerous losses

[pearl31]

just going through the Warren Nolan site...
there's a pretty sizable # of teams that have 4+ q1 wins but have a whole bunch of losses, so apparently aren't being considered for at large because of the losses...
So, what kind of formula do you think the committee will employ in sort of where to draw the line? How do they determine either when a team can no longer be considered and when a team can't be higher than an __ seed?

 

[jncuse]

That is a loaded question. They like quantity of quad 1 wins more than %. I would think you if you are 5-10 in Q1, but have a good Q2 record and no real bad losses you may be OK.ough.

But if you have a very good Q2 record (say over 70%) you are also in great shape.

I think less than .500 in Q1+Q2 is a negative, unless a big proportion of those games are against top quad and elite teams.

 

[pearl31]

so let's take a team like Texas.
5-10 q1 and 2-3 q2. No q4 losses. 17-13 overall record. On the bubble, many have them as "in."

Now compare that to a Georgia.
4-7 q1 and 4-3 q2. No q4 losses. 16-12 overall record. No one (at least in bracketology land) even has them on the bubble anymore.

??

 

[jncuse]

so let's take a team like Texas.
5-10 q1 and 2-3 q2. No q4 losses. 17-13 overall record. On the bubble, many have them as "in."

Now compare that to a Georgia.
4-7 q1 and 4-3 q2. No q4 losses. 16-12 overall record. No one (at least in bracketology land) even has them on the bubble anymore.

??

Only have 2 plausible answers, the larger one being #2:
q3 losses are not irrelevant either and they have 2 of them.
Put a "7" in front of your RPI and you will tend to get overlooked, unless you have something to differentiate yourselves (for us in 2016 it was top 50 wins)

 

[tbonezone]

the quadrant win bands (30-40 teams ?) are just way too wide to be of any practical use.
the idea was sound but lacks the proper gradation to be of any use. i predict business as usual.